UPDATE: FDev have begun looking at this and other issues in 2.1:
https://forums.frontier.co.uk/showthread.php?t=257892
By all means keep your suggestions coming. I will keep this post updated in the hope that they might help.
- Limoncello Lizard:
It would be OK if you could keep the previous couple of rolls, so you could choose the best out of three.
- SP4x:
There are fixed improvements that are incremental with the tier of the upgrade along with corresponding drawbacks e.g. longer jump distance but longer FSD charge time. The upgrade still costs time, credits and loot but your time isn't wasted on the roll of the dice. You know what you're going to get for your investment. If an element of RNG was still desired then perhaps only include this in the "Bonus" feature and then only have it cost credits to spin the "Bonus" roulette wheel.
- SP4x (another good idea):
Hey, maybe have both, spacecraft "Chop-shops" - "you pay's yer money, you takes yer chance sonny" and true specialist spacecraft Engineers - "Speed cost's money Commander, how fast do you want to go?"
- OP (me):
Narrow the percentages for every consecutive roll after the first: bring down the upper limit of negative effects, and increase the lower limit of positive ones; thereby providing a good simulation of how a real-life engineer goes through a constant feedback loop to provide better results. This would still require players to spend a good amount of time farming materials and components to roll the dice, but it would cut down grinding time and RNG sucktacular-ness when rolling said dice.
- Mephane:
Remove commodity requirements from engineer upgrades entirely, and replace the global material storage limit with a per-material limit (of a similar magnitude as the global limit is right now).
- Mephane (another good idea):
What's really needed is either no limit at all (apart from the technical 4 billion limitation), or individual limits. Now 600 per material, that sounds much better.
- Fractal:
Either increase storage capacity OR reduce the list of components by a good 30%. And add possibility to store some commodity outside your ship
- CMDR ColD_ZA:
My solution to engineers would be to lock the upgrades behind rep with the engineers, and make their rep harder to get. That solves the problem of getting the players to experience the new content because they will have to do missions etc to get rep with the engineer... and once you done "grinding" rep, you can purchase the mods for credits; that way the grind is a one off. Also if you pick your special then you loose rep, and have to grind a bit more rep. My suggestion falls in line with the game's current philosophy, whereas the current system feels very tacked on/out of place, in my opinion
- moose666:
A lot of the materials are tiered versions of similar/the same thing, so my suggestion would be: Add a salvage merchant/junk dealer to stations, where you can trade your materials up and down the tiers i.e. 4 worn shield emitters = 1 shield emitter and vice versa.
- Snarfbuckle:
Another idea could be that the CMDR is allowed to set the DESIRED parameters for the job. The cost for the modification will increase exponentially the better the desired mod result. The modifications must be within the engineer rating. Adding surplus materials above the minimum allows for the engineer to test the design more and reduce the chance for a bad roll
Eera:
What it will be nice, is that the equipments upgrade is based on a simulation. Before to proceed to any upgrade, we could ask engineer to use an update simulator. This simulator could show the stats result. If the player think that those stats are acceptable, then, player could ask engineer to proceed to the upgrade and then use materials. If not, player could ask a new simulation, in cost of reputation or credits (less hard to get).
MuetDhiver:
Here is my take what could be improved: Possibility to exchange lower grade stuff for higher grade (e.g. 3 worn shield emiters for 1 shield emiter ), and vice versa (high grade -> low grade). Match missions rewards to pined blueprints materials requirements. Remove negative secondary effects. These really can screw upgrades hard. Make it an only positive "chancy" buff.
Muetdhiver (another good idea):
An elite explorer should get the rep for rank 3 access no question asked with exploration themed engineers. Same for trading and combat rank with the coresponding themed engineers.
Becky:
How about having a few Stations dotted around that accept and resell engineer items as commodities. So people who don't want/need certain items can sell them there in a few central locations; and other people who need them can buy them at a small markup.
InDigital:
What about unlocking grades for Rear Admiral or Duke? Or if you are dangerous, deadly, elite? You know I mean unlocking grade at appropriate rank. Unlocking ability to buy upgrades or to buy materials at appropriate ranks. Again, i mean... ranks could come handy if Engineers could use em. All the spent time for ranking up should get some yield from the game.
Arjin:
Given that the list of materials huge, I would like to see them more concentrated per source, for example, USS - Degraded Emissions, would drop up to 5 specific materials at random but only from that pool of 5. USS-Encoded Emissions would drop a different group of 5 materials from it's pool, and always have a Private Data Cache available to scan which in turn would have its own specific pool of 5 data scans. This would help alleviate the sheer frustration for players in trying to find the materials/components as they would know that a specific source, signal, base, or loot from ships will always drop what they need. Even though this is still random, it is only from a pool of 5 materials per source which greatly reduces the enormous RNG of finding materials.
https://forums.frontier.co.uk/showthread.php?t=257892
By all means keep your suggestions coming. I will keep this post updated in the hope that they might help. Proposed solutions!
- Limoncello Lizard:
It would be OK if you could keep the previous couple of rolls, so you could choose the best out of three.
- SP4x:
There are fixed improvements that are incremental with the tier of the upgrade along with corresponding drawbacks e.g. longer jump distance but longer FSD charge time. The upgrade still costs time, credits and loot but your time isn't wasted on the roll of the dice. You know what you're going to get for your investment. If an element of RNG was still desired then perhaps only include this in the "Bonus" feature and then only have it cost credits to spin the "Bonus" roulette wheel.
- SP4x (another good idea):
Hey, maybe have both, spacecraft "Chop-shops" - "you pay's yer money, you takes yer chance sonny" and true specialist spacecraft Engineers - "Speed cost's money Commander, how fast do you want to go?"
- OP (me):
Narrow the percentages for every consecutive roll after the first: bring down the upper limit of negative effects, and increase the lower limit of positive ones; thereby providing a good simulation of how a real-life engineer goes through a constant feedback loop to provide better results. This would still require players to spend a good amount of time farming materials and components to roll the dice, but it would cut down grinding time and RNG sucktacular-ness when rolling said dice.
- Mephane:
Remove commodity requirements from engineer upgrades entirely, and replace the global material storage limit with a per-material limit (of a similar magnitude as the global limit is right now).
- Mephane (another good idea):
What's really needed is either no limit at all (apart from the technical 4 billion limitation), or individual limits. Now 600 per material, that sounds much better.
- Fractal:
Either increase storage capacity OR reduce the list of components by a good 30%. And add possibility to store some commodity outside your ship
- CMDR ColD_ZA:
My solution to engineers would be to lock the upgrades behind rep with the engineers, and make their rep harder to get. That solves the problem of getting the players to experience the new content because they will have to do missions etc to get rep with the engineer... and once you done "grinding" rep, you can purchase the mods for credits; that way the grind is a one off. Also if you pick your special then you loose rep, and have to grind a bit more rep. My suggestion falls in line with the game's current philosophy, whereas the current system feels very tacked on/out of place, in my opinion
- moose666:
A lot of the materials are tiered versions of similar/the same thing, so my suggestion would be: Add a salvage merchant/junk dealer to stations, where you can trade your materials up and down the tiers i.e. 4 worn shield emitters = 1 shield emitter and vice versa.
- Snarfbuckle:
Another idea could be that the CMDR is allowed to set the DESIRED parameters for the job. The cost for the modification will increase exponentially the better the desired mod result. The modifications must be within the engineer rating. Adding surplus materials above the minimum allows for the engineer to test the design more and reduce the chance for a bad roll
Eera:
What it will be nice, is that the equipments upgrade is based on a simulation. Before to proceed to any upgrade, we could ask engineer to use an update simulator. This simulator could show the stats result. If the player think that those stats are acceptable, then, player could ask engineer to proceed to the upgrade and then use materials. If not, player could ask a new simulation, in cost of reputation or credits (less hard to get).
MuetDhiver:
Here is my take what could be improved: Possibility to exchange lower grade stuff for higher grade (e.g. 3 worn shield emiters for 1 shield emiter ), and vice versa (high grade -> low grade). Match missions rewards to pined blueprints materials requirements. Remove negative secondary effects. These really can screw upgrades hard. Make it an only positive "chancy" buff.
Muetdhiver (another good idea):
An elite explorer should get the rep for rank 3 access no question asked with exploration themed engineers. Same for trading and combat rank with the coresponding themed engineers.
Becky:
How about having a few Stations dotted around that accept and resell engineer items as commodities. So people who don't want/need certain items can sell them there in a few central locations; and other people who need them can buy them at a small markup.
InDigital:
What about unlocking grades for Rear Admiral or Duke? Or if you are dangerous, deadly, elite? You know I mean unlocking grade at appropriate rank. Unlocking ability to buy upgrades or to buy materials at appropriate ranks. Again, i mean... ranks could come handy if Engineers could use em. All the spent time for ranking up should get some yield from the game.
Arjin:
Given that the list of materials huge, I would like to see them more concentrated per source, for example, USS - Degraded Emissions, would drop up to 5 specific materials at random but only from that pool of 5. USS-Encoded Emissions would drop a different group of 5 materials from it's pool, and always have a Private Data Cache available to scan which in turn would have its own specific pool of 5 data scans. This would help alleviate the sheer frustration for players in trying to find the materials/components as they would know that a specific source, signal, base, or loot from ships will always drop what they need. Even though this is still random, it is only from a pool of 5 materials per source which greatly reduces the enormous RNG of finding materials.
Original Post
Howdy folks
My intention is to address only one of the many new features in 2.1: the RNG (random number generator, for non-programmers) behind engineer-driven module upgrades. So here I go...
Firstly it feels very lazy (to me) that exceptionally specialized engineers in the game cannot define, with any modicum of precision, what the effects of a module upgrade will be. If the goal was to aim for realism, I dare say this is a miserable failure. These are supposed to be 30 ultra-specialized engineers in a galaxy with trillions of people, and quite possible billions of them engineers themselves. The exceptionally high quality of the work of these 30 masters of their craft should put them on-par with the likes the people at NASA or CERN here on 2016 Earth, but in this game they seem more like the kind of mechanics you'd find in a run-of-the-mill-backwater-planet's bicycle repair shop.
In the real world when you find yourself in need of a specialized engineer, and said professional asks that you provide the materials to complete the job, he/she will usually mention characteristics like the purity needed, clarity, strength, flexibility, tension, and so on. And if for some reason they're not easy to find, said engineer (the good ones anyway) will either ask you for more resources so he/she can get the source materials up to spec or will point you in the direction of someone who can. Regardless of how it's done, the finished job will rarely have such wildly varying results as the ones produced by the so-called "engineers" in this game. Do you think the Space Shuttle would fly straight if one of its rocket engines produced 10% less thrust than the other?
In the virtual world mechanics like RNG crafting are usually a source of much grief, particularly for those of us with real world lives (families, friends, jobs and careers), since the system can only be defeated by spending spectacularly long amounts of time grinding away in search of good rolls and supplies (the last roll was awful? well too bad, your materials are gone, off to the grind again). What makes the case of Elite Dangerous 2.1 worse, is the amount of upgrades available for such a large amount of modules and ships.
Image, for the sake of argument, that the RNG takes 1 of a maximum of 5 possible numbers every time the dice roll, and that the RNG cannot roll any of the previously selected numbers. This means it would take at most 5! (or 120) rolls of the dice to get the top score. For the sake of simplicity, I'll refer only to the average amount of rolls/materials needed, which is half that: 60.
If faster thrusters were the goal, you're in luck: farm ~60X times the materials needed and you will, after ~60 rolls, get what you want. If you're after two module upgrades... well, your job just became twice the hassle: you'll need ~120X as many materials and just as many rolls to get both modules with top or close to top-notch stats. There are a total of 8 Core Modules of which, I think, only 4 are worth the trouble of upgrading: FSD, Thrusters, Power Plant and Power Distributor (if any of you dear readers has a soft spot for Sensors, Life Support, or Fuel Tanks, please forgive me I meant no offense - and yes, I'm completely ignoring bulk heads, again, sorry).
Speaking of time... after grinding away what little free time you have looking for materials (granted, not all are hard to find), you have to factor in the seconds it takes for the engineer to roll the dice. It's an animation that takes about 3-5 seconds per roll. So for the simple example above, just rolling the dice 60 times would take 3-to-5 minutes of clicking and lightning fast reflexes to not pass up a good roll. There are also upgrades for non-core modules like Shield Generators, Shield Cells, FSD Interdictors and so on, so you can see how this whole process can get very tedious extremely fast.
Now the reality is that the RNG doesn't have a pool of only 5 numbers to pick from, it has a lot more: if you're in luck it's a 1-byte data type (C++ unsigned char) with 255 values in the pool. If FDev was particularly mischievous (which I can't confirm or deny, I haven't seen the code) they could be using a 2-byte data type (C++ short int) with 65,535 possible values in the pool, or worse, a regular sized 4-byte data type (C++ int) which would have 4,294,967,295 values in the pool. As if that wasn't bad enough, and if this works like any other RNG algorithm out there, it can and will roll any of the previously selected numbers. Which means we're talking to N-to-the-power-of-N combinations... now you get to multiply that for number of positive and negative effects listed in the recipe and out to farm we go! Yay! Fun! Not...
In conclusion, this is a horrible system, and while I do not wish to offend anyone at Frontier Developments, I would like to point out that whoever thought this would be a good gameplay mechanic deserves the equivalent of a Razzie award for Game Design.
PS: I do sincerely apologize if anything I wrote here offends you, dear reader. If you would like me to rephrase some of it (hopefully not all of it), please don't hesitate to ask me via private message or through a public reply. I will get to it as soon as I possibly can.
CLARIFICATION: My approach to the RNG mechanics behind the curtain was a bit unfair, to say the least. CMDR Runcible Shaw has correctly pointed out that it might not be as bad as I made it look. Neither of us may know (with any degree of certainty) how things were implemented by FD, but it would be unfair if I didn't even mention it. Here's his excellent explanation of how the RNG system might be implemented.
My intention is to address only one of the many new features in 2.1: the RNG (random number generator, for non-programmers) behind engineer-driven module upgrades. So here I go...
Firstly it feels very lazy (to me) that exceptionally specialized engineers in the game cannot define, with any modicum of precision, what the effects of a module upgrade will be. If the goal was to aim for realism, I dare say this is a miserable failure. These are supposed to be 30 ultra-specialized engineers in a galaxy with trillions of people, and quite possible billions of them engineers themselves. The exceptionally high quality of the work of these 30 masters of their craft should put them on-par with the likes the people at NASA or CERN here on 2016 Earth, but in this game they seem more like the kind of mechanics you'd find in a run-of-the-mill-backwater-planet's bicycle repair shop.
In the real world when you find yourself in need of a specialized engineer, and said professional asks that you provide the materials to complete the job, he/she will usually mention characteristics like the purity needed, clarity, strength, flexibility, tension, and so on. And if for some reason they're not easy to find, said engineer (the good ones anyway) will either ask you for more resources so he/she can get the source materials up to spec or will point you in the direction of someone who can. Regardless of how it's done, the finished job will rarely have such wildly varying results as the ones produced by the so-called "engineers" in this game. Do you think the Space Shuttle would fly straight if one of its rocket engines produced 10% less thrust than the other?
In the virtual world mechanics like RNG crafting are usually a source of much grief, particularly for those of us with real world lives (families, friends, jobs and careers), since the system can only be defeated by spending spectacularly long amounts of time grinding away in search of good rolls and supplies (the last roll was awful? well too bad, your materials are gone, off to the grind again). What makes the case of Elite Dangerous 2.1 worse, is the amount of upgrades available for such a large amount of modules and ships.
Image, for the sake of argument, that the RNG takes 1 of a maximum of 5 possible numbers every time the dice roll, and that the RNG cannot roll any of the previously selected numbers. This means it would take at most 5! (or 120) rolls of the dice to get the top score. For the sake of simplicity, I'll refer only to the average amount of rolls/materials needed, which is half that: 60.
If faster thrusters were the goal, you're in luck: farm ~60X times the materials needed and you will, after ~60 rolls, get what you want. If you're after two module upgrades... well, your job just became twice the hassle: you'll need ~120X as many materials and just as many rolls to get both modules with top or close to top-notch stats. There are a total of 8 Core Modules of which, I think, only 4 are worth the trouble of upgrading: FSD, Thrusters, Power Plant and Power Distributor (if any of you dear readers has a soft spot for Sensors, Life Support, or Fuel Tanks, please forgive me I meant no offense - and yes, I'm completely ignoring bulk heads, again, sorry).
Speaking of time... after grinding away what little free time you have looking for materials (granted, not all are hard to find), you have to factor in the seconds it takes for the engineer to roll the dice. It's an animation that takes about 3-5 seconds per roll. So for the simple example above, just rolling the dice 60 times would take 3-to-5 minutes of clicking and lightning fast reflexes to not pass up a good roll. There are also upgrades for non-core modules like Shield Generators, Shield Cells, FSD Interdictors and so on, so you can see how this whole process can get very tedious extremely fast.
Now the reality is that the RNG doesn't have a pool of only 5 numbers to pick from, it has a lot more: if you're in luck it's a 1-byte data type (C++ unsigned char) with 255 values in the pool. If FDev was particularly mischievous (which I can't confirm or deny, I haven't seen the code) they could be using a 2-byte data type (C++ short int) with 65,535 possible values in the pool, or worse, a regular sized 4-byte data type (C++ int) which would have 4,294,967,295 values in the pool. As if that wasn't bad enough, and if this works like any other RNG algorithm out there, it can and will roll any of the previously selected numbers. Which means we're talking to N-to-the-power-of-N combinations... now you get to multiply that for number of positive and negative effects listed in the recipe and out to farm we go! Yay! Fun! Not...
In conclusion, this is a horrible system, and while I do not wish to offend anyone at Frontier Developments, I would like to point out that whoever thought this would be a good gameplay mechanic deserves the equivalent of a Razzie award for Game Design.
PS: I do sincerely apologize if anything I wrote here offends you, dear reader. If you would like me to rephrase some of it (hopefully not all of it), please don't hesitate to ask me via private message or through a public reply. I will get to it as soon as I possibly can.
CLARIFICATION: My approach to the RNG mechanics behind the curtain was a bit unfair, to say the least. CMDR Runcible Shaw has correctly pointed out that it might not be as bad as I made it look. Neither of us may know (with any degree of certainty) how things were implemented by FD, but it would be unfair if I didn't even mention it. Here's his excellent explanation of how the RNG system might be implemented.
Last edited:
